Title: RE: CA's AutoSys
I think we may be talking about the same
place Terry, aren't we?
Matt Adams
-Original Message-
From: Ball, Terry [mailto:[EMAIL PROTECTED]]
Sent: Friday, October 18, 2002 4:09 PM
To: Multiple recipients of list ORACLE-L
Subject: RE: CA's AutoSys
Oh, I don
Dennis- The export is from a production system and the import is onto a
development/reporting system, so the import time is not very important. I am
using a QUERY parameter to limit the amount of data, but it is still doing a
full table scan to get the rows. I will think about how I can paralleize
Established company in San Antonio Texas needs an Oracle DBA with solid Data
Warehouse
experience to join its IT team.
*Candidates living in Texas will be given proirity however some relocation
assistance
may be available for the right candidate.
PLEASE Do Not send your resume for this position U
Barb,
There's nothing in the STATSPACK report itself that mentions 25% or 16% or
whatever, so your math skills are just fine... :-)
---
It is the YAPP report (obtained by uploading STATSPACK report to
full-functionality YAPP processor within www.oraperf.com) that shows these
breakdowns. The YA
The problem comes down to this:
If there is a '&' in the script, and there is no value already
assigned to it, sqlplus is going to stop processing, and
wait for user input.
I've never found a way around this.
It would be nice if there were some facility such as all
unix shells have to test for
Hello Rick,
I think that if you use EXISTS instead of IN
the optimizer will have to do a FTS on the big table
because he can't use any avaiable index on it.
And it's a big table
While if you use IN and you have an index in the
parent table (the BIG one) this index CAN be used (here,
field1
Wow. That is a perfectly logical and amazingly simple explanation I would never have
thought of. I am in awe.
Jim
Chris,
The answer may be in what occurs below the water. Remember that
RESOURCE is a bit special (here for Oracle 5 compatibility reasons),
that it is one of the two (I thi
Oh, I don't know about that. At one of the places you used to work, they
_finally_ got it to work, but it did need at least one full-time, dedicated
employee to monitor, maintain and massage it.
Terry Ball, DBA
Birch Telecom
Work: 816-300-1335
FAX: 816-300-1800
-Original Message---
Dennis,
I have read about NOT IN. Oracle doc says the following but not conclusive
"In certain circumstances, it is better to use IN rather than EXISTS. In
general, if the selective predicate is in the subquery, then use IN. If the
selective predicate is in the parent query, then use EXISTS. "
Erik - Here is another thought for you. If the table you are exporting is
that large, how long will it take you to import it somewhere? Import can be
several times longer than export, since it is doing inserts rather than just
a query. Tables that take hours to export can take days to import. What
Hi
I have setup dataguard on v8.1.7.4.I have been noticing some times the same
archived log files shipped multiple at standby archival destination like
arch_1_18959.arc
arch_1_18959.arc.ODGship.ODGbad
arch_1_18959.arc.ODGbad
and some of message in data guard logs.
No messages in either oth alert
Dick - Thanks for posting your results. We have a different ERP (Lawson),
and we have been considering whether Resource Management is worth the
effort. Currently all users access Oracle through a single userid, but there
is an option for each user to have a separate login. The price is that you
the
Rick - I would check the EXPLAIN PLAN for both queries, and see whether
Oracle uses a different query plan. Usually the prescription for better SQL
performance is to avoid NOT IN, as the following snippet that I snagged off
the 'net explains.
Use NOT EXISTS in Place of NOT IN
In sub-query statemen
Thank you all for your feedback. I will try again tonight with the following
changes 1) take all rb segments off line except the large one 2) set the
buffer to 1000 (was default before). I will let you know how things went
(or didnt) on Monday morning.
Thanks again.
Erik
-Original Messa
Ed,
The version is 8.1.6 and yes it was with two different usernames. In the
end since we wanted to apply this to PeopleSoft where everyone is the same
userid it was impossible too. Still looking for another solution.
Dick Goulet
Reply Separator
Auth
Hi All,
I have 2 tables
BIG - 100 million records
SMALL - 1 million records.
I want to delete all the records in BIG that are in small.
There is a PK on field1.
Which of the below methods would you choose and why?
DELETE FROM big
WHERE field1 IN (SELECT field1 FROM small);
DELETE FROM big a
W
HELP
(I'm an idiot)
OK, never claimed math was my strong suit.
Why is this 25% of total response time? (Truthfully, I don't really
understand the column heading "of total.."
Here are the top SQL statements ordered by physical reads per execute:
StatementExecutes Physical
Gaja is not active on the list due to his new workload...
... but he receives 'oracle-l updates' whenever we talk ;)
- Kirti
-Original Message-
Sent: Friday, October 18, 2002 10:54 AM
To: Multiple recipients of list ORACLE-L
I agree with Jay. Everytime I use OEM there are 10 or 20
Dick,
Which version of Oracle ?
Was your test with 2 different user ID's ?
Our problem is that all users log on
with the same user id, which seems
to diminish the product's effectiveness.
thanks.
ed
- Original Message -
To: "Multiple recipients of list OR
Has anyone used Oracle's self-study CDROM's to prepare for the
OCA / OCP exams? If so, I would like to know what you think of
them.
Thanks,
Ken Janusz, CPIM
We too make extensive use of the product, though our team does not need to
maintain it. Just the backend database. Seems to work as intended, alerting
the operations staff in case of failures. I have not heard the folks trash
this product. Though, I have had them ask me for JIL scripts, whatever t
We use CA-Workload which I think is the same product packaged
differently. We schedule about 400 jobs each night through it. The
jobs run on NT, VMS, Oracle, and Oracle Express with interdependencies
between jobs on different platforms. I have to say it works pretty
well, but only after about a
Title: RE: Archive files and their Management
ok thanks for the info. The following was MISLEADING:
http://www.orafaq.com/faqdbabr.htm
Section: "How does one put a database into ARCHIVELOG mode?"
"To enable ARCHIVELOG mode, simple change your database startup command script, and bounce the da
Gaja is the new project manager. some of the people also working on the
product are John Beresniewicz and (I believe) James Morle.
Should be a really good product when they are done with it
--- Ruth Gramolini <[EMAIL PROTECTED]> wrote:
> Beth,
> I heard that the new project manager is Gaja. We
I just ran the statspack output through OraPerf as well.
This is truly an amazing resource.
Jared
"Tim Gorman" <[EMAIL PROTECTED]>
Sent by: [EMAIL PROTECTED]
10/18/2002 09:19 AM
Please respond to ORACLE-L
To: Multiple recipients of list ORACLE-L <[EMAIL PROTECTED]>
c
Title: RE: CA's AutoSys
I too
had issues with this product... But, it was several years ago...
Back then it was owned by Platinum and the product ran on NT 3.51... We
were using it to schedule all our Unix batch jobs for the Sales Force Automation
project I was working on... Unfortunatel
Title: Archive files and their Management
The
key to the issue is very simple, Redo Logs (including archived) contain
information that cannot be recovered by other means (unless you capture EVERY
statement issued). As such, they are to be protected at all costs. The vast
majority of unrecove
Kirti,
I've toyed with the product due to a perceived need here. The experiment
was to have two users loggin to the database both running very nasty queries (we
have the GOD of obfuscated SQL here). On the first pass, with no resource
management in place both queries ran successfully in 30 m
I'm not quite sure what you mean by "modify the database startup procedure." You
must be in MOUNT mode to change between archive and noarchive. You don't need to
change the mode each time you startup you database. Once you put a database in
archivelog mode, you are in that mode until you ch
What would concern me is the 99.x% buffer cache hit ratio
along with an apparently huge number of LIO's per second.
This database is spinning its wheels, due to poor SQL
and/or poorly written procedural code.
That's my 30 second analysis. :)
Is this a commercial or custom app?
Jared
"Leo
Beth,
I heard that the new project manager is Gaja. We can expect a better
product from him and his team.
Ruth
- Original Message -
To: "Multiple recipients of list ORACLE-L" <[EMAIL PROTECTED]>
Sent: Friday, October 18, 2002 11:54 AM
I agree with Jay. Everytime I use OEM there are 1
You try using direct=y on the export to see if it speeds things up enough
that you don't get shot down. There are some cases where you can't do this
(I would have to look it up to get them all): For example, if you are using
the query= parameter. But most of the time, you can greatly reduce
If outdoor life is your thing then consider this company in the Wisconsin
Dells, Wisconsin
area that needs an Oracle 11i DBA to join their staff.
This area is well known for its' boating, fishing, camping, etc.
Relocation Assistance is provided.
PLEASE DO NOT send your resume for this position UN
Title: Archive files and their Management
I'm up
for my backup and recovery class here soon, but this conversation has
brought up questions
that I think some of you wouldn't mind discussing
anyway. I see
where implementing archive logs
requires that you modify the database startup
procedur
Dick,
We have been using that product for quite some time now... and it does a
pretty good job of scheduling things. In our environment *all* jobs are
handled by Autosys, and we are not supposed to use 'cron'. We use the
product for scheduling jobs on a variety of h/w platforms( IBM-AIX, HP, Sun,
Title: RE: CA's AutoSys
Run Screaming in the opposite direction.
Product was a semi-disaster at two previous
job sites. Failed to run jobs when scheduled,
failed to detected when dependent jobs failed.
Just plain didn't work well.
Of course this was when Platinum owned the product,
wh
Sorry for my post. mis-read the question completely
regards
naveen
-Original Message-
Sent: Friday, October 18, 2002 8:54 PM
To: Multiple recipients of list ORACLE-L
I'm not sure whether there is another way, but can the following be done?
in the after logon trigger specify: 'SET TRANS
To give credit where credit is due...
I've never bothered to learn how to interpret a STATSPACK report; I've left
that with what little I ever knew about interpreting a BSTAT/ESTAT report.
Rather, I just send the report to the YAPP report post-processor at
www.oraperf.com and it formats everythin
Check the log files in the log directory. If the agent fails to log into
the instance, it will be in the log files. They are found at
$TNS_ADMIN/network/log.
The agent will always wipe out the snmp_ro.ora and services.ora files upon
restarting. It should include entries from oratab and the list
Title: RE: Urgent problem with query in CBO Vs RBO
isn't
this what we pay the maintenance costs for - for them to get this thing
right?
no
excuse for this shoddy bit of work.
-Original Message-From: Johnston, Tim
[mailto:[EMAIL PROTECTED]]Sent: Friday, October 18, 2002 11:54
Hello Listers,
Anyone using Oracle's Database Resource Manager?
Does it really work as outlined in the Docs?
Care to share your experience?
Thanks,
- Kirti
--
Please see the official ORACLE-L FAQ: http://www.orafaq.com
--
Author: Deshpande, Kirti
INET: [EMAIL PROTECTED]
Fat City Networ
Hi,
The snapshot too old error comes when someone has started a query on one
table and at the same time some transaction is going on. So at this time the
select query will read from rollback segment to create a consistent view in
time. Now in what happens select query has started reading the rollba
To All,
I'll admit this is somewhat off-topic, but since the project is going to
handle all scheduled batch jobs that bang on our PeopleSoft Oracle database, I
figured I'd ask if anyone has an opinion preferably from experience using the
product. The Complete A^^holes are here today to do a d
I'm not sure whether there is another way, but can the following be done?
in the after logon trigger specify: 'SET TRANSACTION USE ROLLBACK
SEGMENT..."?
don't know whether this statement is allowed inside a trigger and also
whether it will work.
and also, export commits periodically, the stateme
Title: RE: Urgent problem with query in CBO Vs RBO
"Looks like the lid on the "jar of bugs" is
leaking ... and it made it into 920 code as well."
I think this a pretty common
occurrence at Oracle... I think it has to do with the way they branch
their code streams... Subsequent patches are
I agree with Jay. Everytime I use OEM there are 10 or 20 annoying
little things that make it incredibly frustrating to use.
Jared mentioned a while back that the new product manager for OEM is on
this list. Maybe he could set up something (like a *big* mailbox) to
take suggestions for changes
Scrap OEM. Seriously, we ran into way too many of these little annoying problems. We
decided that we had better things to do with our time. Have you had to re-register all
of your jobs, yet? That one was a real pain.
Jay
>>> [EMAIL PROTECTED] 10/18/02 09:38AM >>>
Hmm, thanks to all who are r
Erik - The problem lies not with your export but with the other
transactions. Basically, export is taking a query of the table when it
starts (like a select * with no where clause). As it is exporting the table,
other applications are making changes to the table. These changes are held
in the rollb
Er, what I meant to say in my last email is to delete the node from OEM and
then rediscover all of the databases on it fresh.
Mike
From: John Weatherman <[EMAIL PROTECTED]>
Reply-To: [EMAIL PROTECTED]
To: Multiple recipients of list ORACLE-L <[EMAIL PROTECTED]>
Subject: RE: OEM can't seem
The only correct resolution to this is to ensure that DML transactions
referencing the tables being backed up, are not run at the same time.
Bringing the big rollback segment online wont buy you anything. You could
delay the inevitable by removing the optimal clauses if set.
Raj
When you do a lsnrctl status is there anything different for the instance that isn't
discovered and the ones that are?
-Original Message-
Sent: Friday, October 18, 2002 9:39 AM
To: Multiple recipients of list ORACLE-L
Hmm, thanks to all who are replying. It looks like the instance i
Has this instance ever been successfully discovered before? I've found that
if I delete it from OEM it will then discover the database.
Mike
From: John Weatherman <[EMAIL PROTECTED]>
Reply-To: [EMAIL PROTECTED]
To: Multiple recipients of list ORACLE-L <[EMAIL PROTECTED]>
Subject: RE: OEM ca
Is this UNIX or Windows?
Where is your oratab file located and what is your ORACLE_HOME directory?
The discovery process looks in your oratab (which can be located anywhere as
long as the ORATAB environment variable is defined) or the Windows registry
to locate all of the ORACLE_HOMEs that are in
Title: SNP acronym?
Saturday Night Paging
-Original Message-From: Markham, Richard
[mailto:[EMAIL PROTECTED]]Sent: Thursday, October 17,
2002 5:29 PMTo: Multiple recipients of list
ORACLE-LSubject: SNP acronym?
What does the background job queue acronym SNP stand for
?
LOL, don't I hear THAT all the time!
-Original Message-
Thomas F
Sent: 18 October 2002 14:24
To: Multiple recipients of list ORACLE-L
this is why I gave up on OEM.
it sometimes discovers things and sometimes not. I just use Dba Studio now.
-Original Message-
[mailto:Reginald.W
John,
Try this
On the client node:
1. Make sure all instances are in oratab & listener.ora.
2. Stop the agent.
3. rm $ORACLE_HOME/network/agent/*.q
4. rm $ORACLE_HOME/network/agent/services.ora
5. rm $ORACLE_HOME/network/admin/snmp_r*.ora
6. start the agent
7. Check the pertinent log files in $OR
Title: RE: Urgent problem with query in CBO Vs RBO
I don't think the reply I sent last night made it to the list.
We eventually traced it to the behavior of _unnest_subquery parameter. Setting it to false solved the problem. This was first reported in 817 and supposed to be fixed in 8173 and
Thanks for your contributions on this Tim (and on everything else you
respond to).
These little tips of how to analyze statspack reports properly all add up
and whilst I did look at the report and I did glean some of you what
suggested I certainly did not pick up all that you spotted
John
George,
Two things jump out together:
The SQL statement with hash value = 3509998681 is
consuming about 25% of the total response-time (i.e.
total processing plus total wait) on the system. This SQL statement is
executing 900 times during the one-hour sample period...
Waits on t
I am performing an export of a set of very large tables. I am not using
CONSISTENT mode. The exports are failing due to snapshot too old errors. The
application uses many small rollback segments, but has one large one for big
transactions. Is it possible to force the export to use the large segment
hi!
then i think this is what you need:
select to_date('01-01-'||round(to_char(sysdate, ''), -1),
'DD-MM-') from dual;
reg.
daniel
Imran Ashraf wrote:
>
> Im running this query:
> select round(add_months(sysdate,500),'')
> from dual;
> This retuns:
> 01-JAN-2044.
>
> I want
Hmm, thanks to all who are replying. It looks like the instance is not
making
it into either snmp_rw.ora, snmp_ro.ora or the services.ora. I have
manually
edited them, but when I restart the agent, it wipes out the manual entries
in
the snmp_ro.ora and the services.ora.
The oratab and listener
this is why I gave up on OEM.
it sometimes discovers things and sometimes not. I just use Dba Studio now.
-Original Message-
[mailto:Reginald.W.Bailey@;jpmorgan.com]
Sent: Friday, October 18, 2002 5:14 AM
To: Multiple recipients of list ORACLE-L
The instance should be listed in the or
I do an rman backup of the archivelogs each time I do a backup of the
database. I use the delete option and delete the archivelogs because I can
restore them from the backup.
HTH,
Ruth
- Original Message -
To: "Multiple recipients of list ORACLE-L" <[EMAIL PROTECTED]>
Sent: Friday, Octobe
Hello list
I use this script to do reorganize - redimension.
It show the use/free space of each table and index. You can change it and then
move/rebuild.
You must change USER and TABLESPACE.
Do not do it ONLINE, at least with big tables. (I do it online and get ORA-0600s).
Connect as the table
Title: Archive files and their Management
theoretically only the archive files after the backup started are needed.
but there might be issues of a corrupted backup, and to guard against that
we have a policy of keeping archive files for one week.
you
should also keep checking your backups b
Hello George
I think that you need to tune the first statement the do the most gets and
the most i/o.
The same statement also access the most rows.
It is fired up about every 5 seconds and is very resource consuming.
I am not a tuning expert but this is my 0.01$ worth.
Yechiel Adar
Mehish
-
Im running this query:
select round(add_months(sysdate,500),'')
from dual;
This retuns:
01-JAN-2044.
I want it to return:
01-JAN-2040.
Thanks
--- Daniel Wisser <[EMAIL PROTECTED]>
wrote:
> hi imran!
>
> don't exactly understand what you want to do, but
> rounding
> e.g. the year work
>Hi,
>
>how do i round dates like this:
>Date = 01-01-1996
>Return = 01-01-2000
>
>or
>
>Date = 01-01-1992
>Return = 01-01-1990
>
>
>Thanks.
>
create or replace function decade_round(p_date in date)
return date
is
d_result date;
begin
d_result := to_date('01/01'||ltrim(round(to_number(
Title: Archive files and their Management
Our policy is to keep archive for the last 2 days
at least.
We are doing daily backup so it gives us the option to
restore from the last backup or the one before
without
restoring archive logs.
Yechiel AdarMehish
- Original Message -
hi imran!
don't exactly understand what you want to do, but rounding
e.g. the year works like that
select round(to_char(sysdate, ''),-2) from dual;
i dont understand what you wanna do with days and months,
but if its always 01-01 then
select '01-01-'||round(to_char(sysdate, ''),-2) from
Hello Viktor
I think that the problem is not with oracle.
If you have the same nls_lang on the client and the
database
oracle will not do translation but will work on the
principal
of GIGO (garbage in, garbage out), meaning what you put
in
is what you get.
You need a program on the client
Hi,
how do i round dates like this:
Date = 01-01-1996
Return = 01-01-2000
or
Date = 01-01-1992
Return = 01-01-1990
Thanks.
__
Do you Yahoo!?
Faith Hill - Exclusive Performances, Videos & More
http://faith.yahoo.com
--
Please see the officia
Title: Archive files and their Management
Greetings Gurus
Just wondering...
Our archive log's directory has grown substantially and space is becoming an issue.
How do you know which archive files is safe to delete?
In other words...
Do you delete all archive files older than the last ba
The instance should be listed in the oratab file, the listener.ora file and
the snmp_rw.ora file. It is automatically added to the snmp_ro.ora file at
agent start up.
Ensure that the Intelligent Agent is running. If any changes need to be
made to any of the previously mentioned files, stop and
I need to estimate disk space sizing requirements for a new database
server. Does anyone have a document that I can send forth to the
developers, asking for inputs from them, about the objects, the growth
patterns, et all? Or any links?
Thanks
Raj
--
Please see the official ORACLE-L FAQ: http://
- Forwarded by Rajesh Rao/CHASE on 10/17/02 05:15 PM -
Rajesh Rao
Yup
The original oracle parameters were called
snapshot_interval
snapshot_processes
or something like that.
In v9, you get a CJQ (Job Queue Coordinator -
obviously some dyslexic at Oracle), and Jnnn's
hth
connor
--- "Johnston, Tim" <[EMAIL PROTECTED]> wrote: >
It's the oracle job queue p
Thanks Mladen and Lee
I am always amazed by the ability and knowledge of the people on this list.
You guys (all the list members) always come up with many options to do
something so if some option does not work you have others.
Yechiel Adar
Mehish
- Original Message -
To: Multiple recipi
You have to be logged on to the machine in order to START the TNS Listener,
in any case.
But to STOP, setting ADMIN_RESTRICTIONS_ doesn't matter;
that parameter only prevents runtime parameter changes (forcing parameter
changes through editing of "listener.ora" and STOP/START or RELOAD).
Only pas
80 matches
Mail list logo
